Envisioning the AI Age: Interview with Yasuko Kosaihira, CEO of Findability Sciences’ Joint Venture with SoftBank Corp.

Findability Sciences Inc., a US-headquartered global provider of AI and big data solutions, established a Japan-based joint venture called Findability Sciences K.K. with SoftBank Corp. (TOKYO: 9434) in 2017. On April 1, 2023, Yasuko Kosaihira took the helm of Findability Sciences K.K. as its new CEO.

Why did she take this challenging new role and what are her thoughts on how AI could transform society? Kosaihira spoke about these topics in this interview.

From back office administration to leading a team of passionate AI engineers

Tell us about your career so far.

I started off at a domestic consulting firm right out of university, and after about three years, I took a job at the former Tohmatsu Consulting, where I worked on M&A and corporate restructuring projects. Gradually, I developed an interest in working for an operating company, which led me to join Fast Retailing. There, I was involved in IT audits and operational audits of its overseas subsidiaries. Later, I joined SoftBank Corp., which was starting to focus on Robotic Process Automation (RPA) at the time, and I was assigned to the Enterprise Business Unit. As I worked on various tasks, I transitioned to a department that oversees investments and alliances, and most recently, I was involved in managing WeWork Japan.

From RPA to AI, your career has evolved quite a bit.

Although I didn't have a background in RPA or AI, I always had an interest in these areas since my auditing days. I heard about technologies that could simplify complex tasks. In auditing, we handled a lot of data, and a lot of the work involved manual checks and cross-referencing. However, there are limitations to the routine checking methods used in audits. I've always thought that AI might be the best way to take advantage of accumulated data to gain new business insights.

In SoftBank's RPA department at the time, we had an instance where we used RPA to stream inventory management information into Pepper-a humanoid robot that many companies were adopting at the time-and we conducted a trial. It's tough for workers to find the products customers want in a large store. Pepper was able to read the product information with the cameras embedded in its head and show product availability on its chest tablet to facilitate purchases. If a product was out of stock, Pepper displayed a QR code to guide customers to an online store. We didn't end up turning this into a service due to some issues, but it was a fun experience.

You were also involved with WeWork Japan.

Yes, I was responsible for streamlining business processes there. Leveraging my previous experiences, I managed and reorganized various functions, including legal, accounting, finance, human resources and sales. Afterward, I was appointed to oversee the operations of around 40 locations across Japan. It was my first time managing a front-facing department that deals directly with customers. Leading a team of about 100 people, addressing daily operational issues, and learning firsthand the importance of middle management gave me new perspectives and instilled new values in me.

After leading 100 people, was it easy to accept a CEO role?

Actually, at first, I wasn't sure if I'd be CEO of Findability Sciences K.K. Having enjoyed working on the front-facing functions of WeWork Japan, I would've been happy to be in charge of the sales department. Then, in January 2023, I had a chance to go to San Francisco on a business trip and I met with Anand Mahurkar, the founder and CEO of Findability Sciences Inc. We were discussing the business outlook when he suddenly said, "So, when should we announce your appointment as CEO of the Japan company?" I was surprised, but things quickly moved forward from there.

Becoming a CEO didn't change my stance toward the company much. Since my consulting days, I've always treated the companies I've been involved with as if they were my own. I've always started by conducting a general check, and then moved on to make improvements. In this case, I saw it as just a step up in the decision-making hierarchy. But when I actually took on the role, I fully felt the scope of my authority and the weight of my decisions.

What was your impression of Findability Sciences from a CEO perspective?

SoftBank embarked on its AI business journey in earnest around 2016, and Findability Sciences K.K. was established as a joint venture then. However, at that time, there was a perception gap between what people thought AI could do and what it was actually capable of. There was a period of disillusionment where we realized that AI wasn't the magic tool we initially thought it was. But during COVID-19, when there was a common understanding that we needed to change the way we work, AI fit the bill, and we've been able to agree on its role in the past couple of years. In that sense, after having struggled for several years with the divergence between market expectations and technology, I think Findability Sciences K.K. is at a turning point right now.

Our team is made up of an extremely talented group of engineers, capable of tooling not only machine learning but also image recognition and deep exploration, and the company is brimming with a curiosity and passion for development. It's very interesting, but because we can build everything from scratch, we tend to work on projects in different markets and themes each time. To effectively utilize such valuable human resources, we need a strategy that efficiently utilizes the strengths of the company. I'm making proposals in that area, drawing on my experience. I feel we have a good relationship where we all can respect each other's areas of expertise and complement one another.

Imagining a Future Envisioned by SF Writer Shin'ichi Hoshi

What are you conscious of when running a Japan-based joint venture of a global company?

Before becoming CEO, I visited two offices in India where development takes place in February. Most of the members are of Indian backgrounds, and I was the only Japanese person. Even though we refer to the country simply as "India," there's a large mix of religions and languages, so we all basically communicated in English. It's truly a model of diversity. Even at dining events, conversations like "I'm in a fasting period right now, so I can only eat these things" came about naturally. I tried to understand the beliefs of about 40 colleagues, but it was too much to grasp. So midway through my mindset changed to one of "there's a variety of customs, and they're all good."

With our headquarters in Boston, and other locations in Los Angeles, India, Dubai, and Japan, time differences are a real issue due to the dispersed nature of our team's working locations. However, as this situation predated the pandemic, our members are very accustomed to working via video conferencing. Even if a certain location can't join a meeting, there's a casual understanding that we can watch a recording later. It doesn't matter where we work, and it doesn't matter whether the camera is on or off. We don't question each other's principles or methods. It's what I'd call "super diversity."

In addition, nearly half of our engineers are women, and they have a major presence. Gender aside, our collaborative projects are moving forward smoothly. India currently has a population of around 1.4 billion and is seen to be a driver of the global economy. As a new set of values are defined in this environment, I believe there's a lot that Japan can learn from India.

Have you made any new resolutions as CEO?

In my new position, I've re-evaluated my strengths and realized I'm the strategic type. If that's the case, then I want to be someone who can draw exciting, interesting roadmaps. I'd be happy if both our customers and employees find the long-term strategies we present to be enticing ones.

So far, I've often felt that quick, short-term results were expected from AI. For example, if we try to recreate the skilled craftsmanship of artisans that would be replaced by AI, and we deem the AI to be useless because it can't surpass the skills of the artisans at this point, AI will never replace them. So we need to think about how to utilize AI for long-term issues, and how to update it so that artisanal skills are properly embedded within AI. We also need to master the new AI technologies that emerge one after another. Given that situation, in my view, we're now at a major turning point, similar to when motor vehicles and the Internet were first introduced, and I want to present a vision for it. This is indeed our starting point.

What's behind your strong commitment to utilizing AI?

I think there are many things in work and life that feel like a bother. While many people try to overcome these things head on, I've always thought, "Isn't there a simpler way to do this?" During my time as a consultant, there were busy periods when I had to analyze data until midnight, and I could feel my mental faculties slowing down, and this impacted my health. There's a limit to what humans can do through grit and effort, so we should delegate what we can to AI. Of course, we need people who can reliably execute on decisions that are made. However, I think we need a certain number of people like me who say, "Wouldn't it be easier if we did it this way?" The world may change from that point on.

Also, I admire the works of the novelist Shin'ichi Hoshi, and his vision of the future had an impact on me. Imagining his cynical and dystopian world, I intend to actively articulate a world that has not been realized by humans by thinking about, for instance, "If we cross this information with that technology, we could do this," and "It would be interesting if we could do this with technology that does not currently exist."

What can we expect in the future from Findability Sciences K.K.?

Strategically, we're targeting industries where significant change is expected, with a focus on life sciences. In the pharmaceutical and medical industries, even with a single disease name or drug name, there are major discrepancies with the naming rules between Japanese compound words and English words, and there are significant hurdles for non-Japanese companies to enter Japan due to local regulations. Findability Sciences K.K. has been working on natural language processing in Japan since 2017 and has a framework that can respond to Japan-specific situations. So we're focusing on this area and making use of our strengths.

We're looking forward to collaborating with SoftBank by harnessing their culture of focus and their sales force. Leveraging a network of thousands of enterprise sales staff, and being able to make proposals to individual companies is a major strength. From the days when we were handling fixed-line communication products, we started dealing with RPA and security products, and our relationship with customers has since evolved. With the addition of AI, we want to focus on a long-term approach to transforming customers' business processes.

AI has the potential not only to lessen the burden of straightforward tasks that humans are repeatedly doing with grit, but also to glean insights from massive amounts of data. Looking ahead, there will be generative AI chats that produce and summarize text, but I wonder about how we can connect them to daily tasks and improvements. I want to make this happen, while possessing the skill to dream like Shin'ichi Hoshi in areas I excel at like redesigning business processes.

(Posted on July 6, 2023, Original article posted on June 16, 2023)
by SoftBank News Editors